Transcript of Helping communities build their own LTE networks | The Stack Overflow Podcast
Helping communities build their own LTE networks JANUARY 10TH, 2022 | 34:52 | E412 SHARE EMBED RECAST SUBSCRIBE EPISODE DETAILS / TRANSCRIPT Esther Jang Each of those sites, the way we've set them up now with our current cellular base stations, the radios that we're using can support technically 96 SIM card users on each network. And that's a significant number, especially because like each SIM card can support a whole little Wi Fi hotspot, or a little home Wi Fi access network, and that would support potentially a whole household. So in addition to using like a SIM card plugged into a mobile phone, you could have like a fixed wireless access device powering Wi Fi in your home, like times 96.
[intro music]
Ben Popper Couchbase Capella database as a service is flexible, full-featured and fully managed. With built-in access via key value SQL and full-text search. It's blazing fast, yet surprisingly affordable. Visit couchbase.com/stackoverflow to try Capella today for free. No credit card required.
BP Hello everybody, welcome back to the Stack Overflow Podcast, a place to talk about all things, software and technology. I am Ben Popper, the director of content here at Stack Overflow, joined as I often am by my two wonderful co-hosts, Ryan Donovan and Cassidy Williams, hello!
Cassidy Williams Hello!
Ryan Donovan Hey Ben!
BP Stronger Seattle contingent now that Cassidy is here, we've really got some Seattle in the house. Today, we are going to be talking about a very cool initiative, what we're talking about is something that I remember reading quite a bit about and writing about when I was at The Verge, which is to what degree you know, communities can create their own internet access, whether that's a mesh network, or whether they're doing their own high speed broadband, or in this case, trying to help communities create their own improved access to LTE, which has become so important as everybody's transitioning to sort of remote and diffuse work. So a super interesting project that's going on. We have three great guests today, Matt, Chris, and Esther. So to the three of you, welcome to the show. Why don't you introduce yourselves and let people know a little bit about who you are?
EJ Hi, my name is Esther, I'm a sixth year graduate student in UDUB computer science. And I've been working on this topic, researching this topic called community networks for the whole time I've been here.
Chris Webb Hi, my name is Chris Webb. I am a professor in the STEM plus V department at Seattle Central College. I'm also here representing the Black Brilliance research project in Seattle. And I am the team leader, and the research lead for the Internet Access and Digital Equity Research Team.
BP Very cool. And Matt, tell the folks who you are.
Matt Johnson Hi everyone. My name is Matt. I'm also a graduate student at the University of Washington, working in community networks, and in particular, this kind of new and different, subtly different concept of community cellular networks.
MP What is the genesis of this project? How did it begin? Who sort of helped to get it off the ground? And what was the need initially that, you know, it was meant to address and maybe how that's evolved over the last few years?
EJ So I guess the thing to emphasize is that all of us in our different contexts, like many different organizations, around the city and around the country have been working on digital equity for a while. And so this project represents kind of an interesting intersection of many of these initiatives and projects in kind of a critical moment. So during the COVID-19 pandemic, everyone has talked about how digital inequities that were existing already in Internet access and digital technology access have been amplified. There's, you know, more conversation about infrastructure disparities between neighborhoods, for example, how in South Seattle, people have less access to high speed Fibre, less access to affordable high speed broadband, people are bandwidth constrained when they're trying to work from home, and they have kids trying to go to school online. And so now there's this conversation about it. But this is a problem that has existed for you know, as long as the internet has existed, these disparities in infrastructure. And so on my side, me and Matt are in a research lab that works on essentially technology for development. We have been working on the problem of Internet access mostly in rural remote areas for many years. So we have all this kind of knowledge and expertise about open source cellular networks. That's kind of what we work on. And we have a great little community of practice, not just in our lab, but globally among people who work on open source cell networks that can be deployed in a community owned and operated way we call that community networks in rural remote villages around the world. Before the start of the pandemic, even our lab had been interested in doing this kind of work in the local area in the Puget Sound. Looking are there Internet access disparities here and wanting to look at whether we could improve people's public access to the internet, in our local communities and maybe lower income neighborhoods. We had started this project not anticipating the pandemic. And then when the pandemic hit, we had applied to a bunch of funding sources that came up for digital equity. And so we started basically promoting this Seattle Community Network project, which was now funded by King County to deploy a few sites in like particularly low income neighborhoods of South King County and South Seattle. And now we're also funded by the City of Seattle IT, but we were always looking to see well, our technology is still kind of experimental, we work on like open source cell networks that are not supported by a company or by like a consistent stream of funds and that sort of thing. So is this this model of open source, community owned and operated networks a good solution for improving public connectivity and access. Throughout the pandemic, we have basically accrued partners who are working on the same goals, and who hear about our project, and we hear about their work. And so we exchange ideas and contacts. And so through the course of the project, we met Chris from Black Brilliance Research, and a number of other partners, including API Chaya, another nonprofit that has a project called Wi Fi As A Lifeline. And so we have, we have all these partners now, including the Tacoma cooperative network as well. So we're working in Tacoma, we're working with BBR here. And in Tacoma, actually, they have like quite a wide reach of community. And we're doing this kind of as a solidarity movement. It's like mutual aid, where we really want to tap into our community partners and resources, all help each other out to build these networks.
RD I was interested, I didn't know you could run your own cellular network. How do you open up the LTE stack so other people can run their own network?
MJ Yeah, it's actually, cell networks in general, are a really interesting kind of parallel branch of technology that's developed, like I think a lot of end users and people at home are used to this idea of like, oh, yeah, I go to the store, and I buy a Wi Fi access point. And I bring it home, and I set it up, and it's mine. And cellular networks are this big foreign thing that, you know, Verizon sets up and AT&T sets up and like how could little old me like build a cell network. And so it's really interesting is over the last 10 years, like there's been a real shift and like the way cell networks themselves are built. Modern cellular networks, starting with LTE and continuing with 5g are actually all based on IP connectivity at the lowest layers. So they actually look a lot more like Wi Fi than they used to. And then in parallel to that, there's been a big movement in open source infrastructure for managing cellular networks. So there's different projects, there's a project called free 5g core, there's a project called magma SRS LTE, and there's a bunch, we in particular, use a project called open five GS. And we settled on that because it offers both a 4g and the 5g core. So we're able to serve users with LTE today, and also have a path forward towards like in parallel, also serving users with 5g with that software stack. On sort of a parallel like piece of this is Spectrum. And so specifically, Wi Fi is really easy to deploy because it uses unlicensed Spectrum. And traditionally, cell networks have only been available in licensed bands. And so you would have to go and get a Spectrum license. And this isn't something that you just go down to like the corner store and buy, you know, you interact with the FCC, and like it's sort of handshakes and business agreements and this really high barrier to entry. But what's changed in the last five years is there's this new spectrum technology in the United States called CBRS, the Citizens Broadband Radio Service. And basically, it's an API for getting access to Spectrum and spectrum that's appropriate for running cell network technologies. So the networks that we're deploying in Seattle use CBRs Spectrum. And then these open source LTE software stacks for running the management layer, the network, and then just off the shelf radios. And so it's actually a possible now to build your own cell network, which is super cool.
BP How come the rapacious capitalists who usually bid on Spectrum didn't gobble that up? What's this nice carve out for Citizens Radio? How did that happen?
MJ Yeah, so I won't pretend to be a policy expert here. So you know, you can speculate, I guess, like so part of it is that at the time CBRS in some ways, it's kind of an experimental band, but actually has this concept of priority access and tiered access. So the band previously was actually used for naval radar. And so the the US Navy had kind of control over the whole band for the entire domain that the FCC managed. And you can imagine it doesn't make tons of sense for the US Navy to be sitting on a bunch To spectrum and Nebraska, and so CBRS, like the idea was, okay, can we share the spectrum in a way that protects the priority users, but like, still allows general access. So the system that they came up with is actually three tiers of access. So there's a priority tier, that's existing operators, so the Navy, and there's some satellite stuff in this band, too. And then the secondary tier, which was priority access. And so the major carriers actually bought most of the priority access licenses, and they went for lots of money in major urban areas. But one of the things that's that, from a policy perspective is really great is that the band was not open for priority access across the whole band. So there's 150 megahertz of spectrum, and only a subset of that was sold for priority access. So there's always a slice that's available for the last tier, which they called General Access. So that's lower power users or users that are able to tolerate like less certainty in the spectrum being available.
RD And you said, it's to get access to kind of this common Spectrum, it's an API call. Is there any sort of, you know, filtering, or reserving? Or do you have to, like, put $1 down to show your theories about it?
MJ Yeah, so it's mediated, there's this idea of a Spectrum access system. And so you'd have to register for an account with a Spectrum access system. Depending on the specifics of the radio that you're trying to operate, you may or may not need, essentially a certification that you can get by attending an online webinar, basically, that certifies that, hey, I know that I'm putting this radio up in a way that I'm not misrepresenting the way the radio works to the Spectrum access system.
EJ When Matt was talking about how much of the Spectrum is like auctioned off versus available for general access, it's actually a significant chunk. It's not just like a tiny band, out of the 150 megahertz, it's half. So only 70 megahertz of that is auctioned off. And then the rest is available, which is significant, that makes a difference. It means that our devices are still okay, even after Verizon starts deploying their gear in the same locations. At first, we were very afraid that oh, the arrival of the telecoms would just kick off all of our gear immediately.
CW I won't speak to the Spectrum piece because there's a bit I wanted to say about the your earlier question Esther answered about community and why we're here, why the Black Brilliance Research Project is here, and how we came to join forces with the University of Washington. So the black noise research project started in 2020. It started out as a result of the Black Lives Matter campaign and protests in summer of 2020. The city of Seattle decided that they would actually allocate money for participatory budgeting process here to the tune of $30 million. $3 million was allocated for us to begin a research project study with community needs here in Seattle, the black community and the bipoc community, as well. And so I was brought on in September of 2020, to lead the digital equity and the Internet access research team. During the process of my research that fall, we discovered the Detroit Community Technology Project, while you all may or may not be familiar with which is an organization based in Detroit, that is doing and has been doing and pioneering along with several other organizations like NYC mesh, these types of community networks, they also developed what's called a Digital Stewardship Program, which is a program to train people, members of the community to be able to not only deploy these kinds of technologies, but also to teach other members of the community about the technology, about information and digital literacy, about the various things that can be done with these networks, and provide access simultaneously. So it was then, it was during that fall that we were introduced to Esther and another colleague of ours, in Tacoma, Emma Slager. And from there, we started having these conversations, hey, y'all are you all have a great solution is very similar to the one that is being implemented in Detroit. Why don't we consider partnering? And that's what happened. So in January, we got together, we all started working on this. We work together and develop the digital stewards curriculum here in Seattle. We're working on a youth digital stewards curriculum and a cohort right now that we'll be launching next week. That's how we kind of all came together. In terms of the work that we're doing in the community and the importance of it. You know, part of our research also was examining 20 years doing secondary analysis on 20 years worth of what are called here the technology access and adoption studies conducted by the city Seattle. Through that secondary analysis process, we ended up working closely with the Seattle IT department and David Keys and his colleagues. And we realized that, that the picture that the City of Seattle had of community needs of community access, stood some room for improvement in their data. So we decided to partner with them as well, to conduct a digital census. And with the knowledge that we had from our research, and our research included over 100 individuals and organizations. Over the course of six months, we were able to identify where the need was that aligned with what the data that the University of Washington had in our partnership was born. So now I'll let Esther and Matt speak to the number of installations that we have, I think we're heading up into six installations where we have a number of other here, just here in Seattle, we have a number of other partners that work with us as well, including Seattle Public Schools.
BP Tell us a little bit about where it's deployed, how many you have, and I guess, you know, six or seven installations? How many people can sort of use that? And what kind of usage are you seeing these days?
EJ So what we sort of initially have funding for is five sites. And then there have been some additional sites that have been funded by the Internet Society. So our first site was actually in Tacoma partnership with the Tacoma cooperative network. Then we have the Filipino Community Center in South Seattle, we have Garfield and Franklin High schools, which are funded by the City of Seattle, we have another potential site in Kent, but we are blocked on some technical hurdles there, where we are looking for a source of upstream internet connectivity, we call that backhaul. And the only current one available because of the relative lack of Fibre infrastructure in Kent, is a relatively expensive connection, that would have to be wired into the church. So it's difficult for us as a small nonprofit slash research lab to actually fund that ongoing expense. So we're looking for different solutions. And we have another site in Tacoma that is not quite ready. Some of our sites, including Franklin High School, and this other site in Tacoma, are backhauled through the University of Washington network, which they donated some bandwidth to us. So we are currently working on setting that up with UDUB IT. And the Tacoma site will be turned on. As soon as it's ready. Each of those sites the way we've set them up now, with our current cellular base stations, the radios that we're using, can support technically 96 SIM card users on each network. And that's a significant number, especially because, like, each SIM card can support a whole little Wi Fi hotspot, or a little home Wi Fi access network. And that would support potentially a whole household. So in addition to using like a SIM card plugged into a mobile phone, you could have like a fixed wireless access device powering Wi Fi in your home, like times 96, we are limited a bit by the amount of total internet capacity that you can get at one of these locations, that one of the cell sites. And so those 96 users would be sharing, for example, a gigabit internet link backhaul link through our network at each site. But depending on like the actual number of users we get, and their their usage patterns throughout the day, that sharing could could work out alright. Right now we're still kind of in the testing phase. Since we don't have a large number of users on any given site yet, we're looking to see how the performance continues to function or degrades as we add more users on each site.
RD So you guys are doing more than just the software, you guys are actually building the devices, installing them, setting up the back, what'd you call it? The backflow connection?
MJ Backhaul.
RD Backhaul. Do you all have other sites in mind?
EJ Yeah, I mean, we are continuously talking to community organizations doing outreach. Right now. We're actually, you know, as a relatively small organization with a number of volunteers, but also just kind of, you know, many graduate students and other and undergraduate students who are busy with other stuff. We are trying to figure out how to scale effectively, while not incurring a huge amount of technical debt. So we're setting up monitoring infrastructure. Right now we're trying to make sure all of our software is stable, trying to figure out good testing and deployment procedures for all of our equipments. And so right now we're mostly focusing on getting the sites that we that we have on deck super stable, and then after that, like, for example, we're talking to the Tacoma Cultural Center, which is another community organization that we know through one of our digital stewards students, and a couple of other community orgs, who have expressed interest, but it takes a while to actually develop the relationships, you need to build those networks. Since they are so physically involved, you need root access on the space, you need to confirm that the space is actually owned by the organization, and that they're allowed to do certain things to the building, we may have to like install conduit for power and things like that. And all of that requires contractors and labor. And then of course, there's even just getting the agreements lined up, like the MOUs for the different organizations involved. And so it is quite a long process, a few months probably.
MJ One interesting thing about cell networks, specifically in this technology space is that because of the way that cell networks work, they're really good at having an access point that's been shared between a large number of users. And so we've really embraced this idea of finding anchor institutions and community partners. And then looking at, hey, we can deploy a cell effectively a small cell tower via an anchor institution, and then having that serve a large number of people in like physical proximity to that institution. This is a little bit different than the way that Wi Fi works, where Wi Fi starts to break down, as soon as you have more than, like 10 users on an access point, like at a low technical level, the way that contention is managed is very different. So this is like one interesting place where the social piece of it, so like the way that you go about building infrastructure in place is impacting the technology and the technology is impacting the social piece of it. So there's this interplay, like it's hard to disentangle the two.
CW Yeah. And I'd like to add to because when you ask the question about what's planned, also, it's important to understand the Black Brilliance Research Project, as well, I'm not only involved in the planning discussions, for the work that we're doing in collaboration, but also in leading discussions for planning that we're doing outside of the city of Seattle. So right now, Tacoma, for instance, we've been in discussions with them for the last six months. But in the past two months, we've actually had the mayor, we've had the workforce development directors, the Tacoma Public Schools, and a number of other organizations in Tacoma that work on workforce development, that have joined to partner with us, and that we are now in the planning phase to actually expand additional networks in Tacoma come with beyond the two that are already in operation. Other organizations also are attempting to begin to plan and make their own plans for expanding these networks. And that's a big part of what the goal was for our partnership together and what Esther's goals were, for working on his project and making sure that it's replicable, and making sure that there are support for other organizations to be able to take on this work and expand it.
EJ Yeah, I just wanted to emphasize that, as much as this is a technical project, it requires a lot of labor that is not technical, just like a huge amount of labor of various kinds, including, especially community relationship building. Because our core mission is to serve people who are under connected, or who have trouble accessing the internet for various reasons, including not having familiarity with devices or being more elderly or having different access needs. We are doing activities like community outreach, and educational workshops, we're planning to have sort of help staff that will not just cover, you know, help with network connectivity, but also other digital technology need, like access needs and problems, especially through some of our digital stewards trained through the Black Brilliance Research Project, who already had these sort of projects and, and goals before the pandemic. And so we're trying to be as community focused as possible in terms of the things that we offer related to technology. And that's really important as a core value of the project.
CW So outside of the hardware and the education of it, and being able to just build everything from the ground up. What do you find to be the biggest barrier for success when implementing one of these towers or implementing a certain program in a given area?
EJ So, from my perspective, Matt, and I have a very weird perspective in that we are computer science graduate students who specifically study the technology. So the technical side for us is not the hardest problem. There are also certain benefits we get like for example, we are familiar enough with the technology that even though it's expensive, like we know what radios to buy, we know what the technical needs are for the equipment. And so we can like cut costs and make things as accessible as possible on the technology side, we know you know where to source things. I'm not minimizing at all the barrier to entry here, because it is significant, each of these radios is about $4,000. And then on top of that, the Spectrum access service, you have to put down a $5,000 deposit to enroll, for example, in the one that we have, and then there's like a license for professionally installing CBRS devices, that is another $600. On top of that. There are significant significant costs here, some of those are covered by our research budget, which is a crazy privilege. But on the other hand, that's, you know, it's a fairly reasonable amount of money that you can put down for a site, you know, as long as you can, like, get together the physical resources that is eminently possible. What's most difficult, in my opinion, is getting together the social infrastructure, having enough social cohesiveness, to actually have people who are in charge of maintenance, people who have good communication with each other and can tell each other when there are problems. People who can ask each other for help. Training people in local neighborhoods has been like a really big, important goal for us, but also really challenging. Like, we need to find people who are available, willing to learn, very persistent. And also there are challenges with, you know, we don't have a way right now to use this network to directly bring in income and pay people. So we are running off of grants. And we're also, you know, trying to reach populations where people may have to have multiple jobs to support themselves, they may have to support families, so they may not have time. And so the time and the resources, in terms of people are a huge challenge in this network.
CW The political piece of this thing, we currently have a few grants that we are working on with multi year grants with the University of Washington. However, the key thing, the social infrastructure that Esther reference, is really important and really complex, when we're dealing with our communities that are usually BIPOC communities, usually some of our immigrant communities, multi generational households, where everyone is required to work. And even with the youth where they have to have a job after school to be able to help support their families and not having that certainty. I think there's definitely the demand, there's definitely an interest, I know there is, in the community, for people to get to learn these technologies, learn these skills, get these get the certifications, as well to be able to build this infrastructure. The problem is, what is the trade off? You know, what's the cost structure for participation? And that's been a big challenge dealing with politicians dealing with, you know, policymakers and trying to get the resources.
BP Yeah, I mean, it's interesting to hear you talk about how grounded in hardware and local, you know, buildings, this is and training and community, it makes me think about, you know, the big infrastructure bill that just passed and what infrastructure means, you know, in the modern world. Certainly, you know, connectivity feels like one of the things that should be in there, and to be able to provide this for the community is something ideally, right, you could apply for a grant or apply for a slice of, you know, federal funds that are meant to help modernize and provide, you know, in some cases, equity among communities. I wanted to ask a little bit about on the ground, what are people's experience of this? Do they understand it right away? Like, can they just come up with a phone and hop on a network? Do they need a certain SIM card? Like, for the end user, what's the experience like, and, you know, having interact with them, I guess what gives you hope and motivation, based on people's, you know, ability to leverage these community networks?
MJ Yeah, that's a great question. And it actually ties back to what I was going to say in response to Cassidy's question too. Which I think, for me, one of the biggest challenges and that crosses, again, both the technical and the social is how do you make the networks visible to people and understandable? And where if you look at cell network technology has been designed to basically hide what's going on. And that's a good thing, like abstraction is a good thing as people maybe who are familiar with software, the software development practice, have experience. But when you're trying to look at empowering people to understand their network, and like make community based decisions about what's going on, you know, there's a trade off there about how much do you hide in for the sake of simplicity, versus how much do you expose for the sake of understanding, there's I think that's both an opportunity and a challenge and a place where there's a lot of room to explore new things and challenge maybe some of the assumptions that are baked into the technology. So like right now, we can operate where yes, we can give people a SIM card and it just works. But then there's a question of if that's what you want.
EJ As the person who onboards people to to the network currently, I definitely don't think that you can just give people a SIM card. There are only certain phone models right now that work with CBRS, the newer phone models specifically. So since we are targeting lower income, marginalized demographics, we generally give them the devices. And that's for right now while the network is targeted at them. We have to give them the SIM cards. And, you know, working with an open source stack where it's not, you know, necessarily production ready, we have some challenges like on certain phone models, you need to put in a certain access point name, just like you do with certain telecom carriers. And so all of that requires some onboarding, the fixed wireless version of the network, where essentially, you have something like a stationary hotspot, or receiver that a SIM card goes in, like that needs to be placed in a window pointed kind of in the direction of the cell tower itself. So it's helpful for us to have the cell tower be as visible as possible. Also, it increases a sense of ownership to have this visibility, like this is our community cellular network, we can actually decorate the antenna, we haven't done this yet. We plan to. We can decorate the antennas and the equipment to be more symbols of community ownership. So I think it's important for us to have like robust training materials and onboarding materials as much as possible, but also to increase this visibility of the infrastructure as much as possible. And in a lot of our outreach and training. We teach people, hey, like, can you spot the cell towers on your street? Can you spot the Fibre access points on the power lines, the places where your home Fibre can be hooked into. And yeah, like making infrastructure visible is really important for us.
[music]
BP Alright, thank you so much for coming on. This was really wonderful to learn about, I don't think I had any idea that this was possible before I got the pitch. I still not sure I fully understand how it all works. But I'm glad that there's Spectrum up for grabs and open source networks are possible. As I do at the end of every episode, I'll shout out the winner of a lifeboat badge someone who came on Stack Overflow, and helped save a question from the dustbin of history. Awarded November 5th, to drewdoormen: a runtime error addition of an unsigned offset. Uh oh. It overflowed. We know about that. Thanks drew for the help. And if you've had some runtime errors or some overflows, you could check out the answer in the show notes maybe to help you too. I am Ben Oopper. I'm the director of content here at Stack Overflow. You can always find me on Twitter @BenPopper. And you can always reach us podcast@stackoverflow.com.
RD I'm Ryan Donovan, I'm a content marketer here at Stack Overflow. I edit the blog, newsletter. I am secretly on Twitter @RThorDonovan and if you have a good blog post idea, you can email me at pitches@stackoverflow.com.
CW My name is Cassidy Williams, I'm director of developer experience at Netlify. You can find me @cassidoo on most things.
MJ I'm Matt, I'm a graduate student at the University of Washington, hopefully graduating soon as PhDs go. And you can find me again on most things at @matt9j.
EJ I'm Esther Jang, PhD at UW computer science. I'm going to graduate, maybe not soon. [Cassidy laughs]
CW Someday!
EJ The website for this project is actually seattlecommunitynetwork.org The nonprofit that is running it is the Local Connectivity Lab. You can find us @SeattleCommNet. My personal handle is at @infrared_ether.
CW My name is Chris Webb. I am with the Black Brilliance Research Project and Seattle Central Central College. You can reach us at inclusivedatasolutions.com. We'd love to hear from you. So please reach out
[outro music]